The study was conducted in two mountain regions in the Western Carpathians, in the central part of Slovakia: Štiavnické vrchy Mts. and Poľana Mts. (Fig. 1). Štiavnické vrchy Mts. (48°12'–48°35'N, 18°32'–19°05'E) occupy a geographical territory of approximately 776.3 km², extending from the hills to the submontane belt. Poľana Mts. (48°35'–48°44'N, 19°18'–19°38'E) represent the highest volcano mountain range in Slovakia, with an area of approximately 183.0 km². The study areas are typical volcanic mountains with a uniform geological substrate. Abiotic conditions led to the existence of diverse vegetation on both mountains. The dominant meadow communities include submontane mesophilic mowing meadows (Arrhenatherion elatioris union) and pastures (Cynosurion cristate union). Wet meadows of submontane and mountain areas (Calthion palustris union), which belong to regionally important communities, are also represented. The geological substrate of the studied sites consists of andesites. Soil types are represented by cambizems and ranchers, ranging from weathered acidic to neutral rocks. These study areas are moderately cold and very humid, with air temperatures ranging from 12 to 16 °C in July. The average annual precipitation is 800–900 mm (Landscape Atlas of the Slovak Republic 2002).

Figure 1.

Location of study areas A Štiavnické vrchy Mts B Poľana Mts.

The vegetation survey was carried out during the growing seasons of 2017 and 2018 at 30 sites. The study sites occurred at similar altitudes, ranging from 481 to 767 m a.s.l., with similar local abiotic conditions. The sites represented 3 ba­sic types of management with more than 10 years of history (based on informa­tion from landlords), with low and medium intensities: mown meadows (hereinafter MGM, mowed once a year, usually at the end of May), grazing meadows (hereinafter MGP, sea­sonal pastures, or fences) and abandoned meadows (hereinafter MGA). Most of the sites have been under the same management type even for a longer time (according to historical mapping from 1957–1971). Each type of management was represented by 10 localities. Typically, there were three immediately adjacent sites, each representing one of the types of management. From a phytosociological point of view, the studied communities can be classified into the class Molinio-Arrhenatheretea Tx. 1937. The coverage of individual vascular plant species was evaluated following the Zurich-Montpellier School of Phytosociology (Braun-Blanquet 1964), using a nine-point scale of coverage and abundance (Westhoff and van der Maarel 1973), on 30 areas of 16 m². The nomenclature of plant taxa is given in the sense of Marhold and Hindák (1998). During the field survey, the altitude was recorded using a GPS device. The slope data were drawn from the publicly accessible database of the National Agricultural and Food Centre. The original interval scale was replaced by mean percentage values for each slope interval.

To characterise the basic physical and chemical properties of the soil, soil samples were taken to a depth of 10 cm from three randomly selected locations in each area at the time of the vegetation survey. Individual samples were combined into a single sample before analysis. The analyses were carried out according to Hrivnáková et al. (2011). For a detailed description of the methods of analysis, see Diviaková et al. (2021). The basic characteristics of the study sites are listed in Table 1.

We selected a set of 10 major PFTs (31 trait attributes) that were supposed to be ecologically meaningful with respect to the studied management types and that we expected to be affected by management. These PFTs included the following PFTs: seed mass, leaf persistence, reproduction type, plant height, duration of flowering, plant lifespan, plant growth form, life strategy, leaf distribution along the stem, and forbs/graminoids. Each plant species was graded for each trait according to the attributes listed in Table 2. The attributes of individual PFTs were evaluated based on 2 databases: the PLADIAS Database of the Czech Flora and Vegetation (Chytrý et al. 2021) and the LEDA Traitbase (Kleyer et al. 2008). All the PFTs monitored attributes were evaluated for each type of management. They were calculated from the averages of the values, weighted using the coverage of each species of vascular plant present at the sites where the sum of all categories was 100%. Differences in basic environmental characteristics among the three management types were investigated using the Kruskal-Wallis test, as variances were not homogeneous according to the Bartlett test, which was performed prior to analysis. The relationships between the proportions of functional groups of grassland plants and the management regime were summarised using redundancy analysis (RDA) with centred response data. The functional category proportion data were square-root transformed prior to analysis. The significance of the relationship was tested with a Monte Carlo permutation test (999 permutations). All analyses were performed in R v.4.1.2 (R Core Team 2021) using the libraries vegan (Oksanen et al. 2013) and pgirmess (Giraudoux et al. 2018).

Compared with abandoned sites, managed sites were characterised by greater coverages of some attributes, e.g., within PFTs leaf distribution along the stem and plant height. Leaf distribution and height are important traits for plant competitive ability and persistence (Drobnik et al. 2011), ergo for plant performance under different management regimes (e.g. Garnier et al. 2007).

We found that rosette species were more successful under the two main management regimes (MGM and MGP). Erect competitor species with leaves distributed regularly along the stem were more likely to occur with higher coverage at abandoned sites (MGA). Similarly, in experimental scenarios of changes in land use, the abandonment or decreasing frequency/intensity of mowing and grazing led to decreases in rosette species and non-branched growth forms, whereas the coverages of taller species and species with leaves distributed regularly along the stem increased (Weiss and Jeltsch 2015), indicating a shift towards stronger above-ground competition (Römermann et al. 2008).

The higher proportion of rosette species, in our study, e.g., Taraxacum sect. Ruderalia, Bellis perennis, Leontodon hispidus, Plantago media) due to vertical defoliation is the main response to grazing and mowing management (McIntyre et al. 1995; Dupré and Diekmann 2001; Klimešová et al. 2008). These species have lower stature, buds closer to the ground, and horizontal leaf orientation and are able to resist grazers (Noy-Meir et al. 1989; Hadar et al. 1999). The rosette species represent an escape strategy that allows plants to survive disturbances and exploit newly available spatial niches (Grime 2001). However, this issue is complicated because, in addition to the form of leaf distribution, other mechanisms may be involved in greater resistance to defoliation. Among those, e.g., low palatability (Perez-Harguindeguy et al. 2003) depending on chemical composition traits (Pontes et al. 2015), mechanisms related to phenology and dormancy (McIntyre et al. 1999), or physical defence, such as spines (Díaz et al. 2001), may be important.

PFT plant height is the trait most frequently used to assess species response to management (Klimešová et al. 2008). Height is important for competitive performance and the acquisition of carbon and is a fundamental functional trait of plants (Westoby et al. 2002). In our study, tall species (> 0.6 m) prevailed (in terms of coverage) at abandoned sites, whereas at managed sites, species of medium height (0.3–0.6 m) or small species (< 0.3 m) showed higher coverage values (cf. Peco et al. 2005; Rupprecht et al. 2016). This pattern could be explained by the higher concentrations of nitrogen observed at the abandoned sites in our study. At abandoned sites, the proportion of tall species usually tends to increase as litter accumulation increases nutrient availability and hinders seedling recruitment (Huhta et al. 2001; Rosenthal 2010). One of the main environmental filters of most plant communities is light (Crawley 1997), and increased competition for light (i.e. avoiding shade) under abandonment (Lepš1999), especially under dense overgrowth, favours tall plants (Westoby et al. 2002; Neuenkamp et al. 2016), such as light competitors (Prévosto et al. 2011). In contrast, species with shorter heights in abandoned localities are the most susceptible to a lack of suitable microsites (Hautier et al. 2009) because of the absence of grazing animals or mowing.

In our study, the abandoned sites had greater coverage of phanerophytes than did the managed sites. Even grazed (MGP) sites showed greater phanerophyte coverage than mowed (MGM) ones probably because they remained in the area due to selective grazing (Dupré and Diekmann 2001). Phanerophytes are unpalatable to cattle and sheep because of their woody stems and the occurrence of spines (e.g., Prunus spinosa and Rosa canina). The presence of phanerophyte species and their adequate control by grazing, which prevents reforestation, can lead to greater structural or physiognomic heterogeneity in grasslands (Kun et al. 2024). In abandoned sites, succession leads logically to afforestation. The most visible consequence of land abandonment is the colonisation of previously opened land by phanerophytes, which are often of high biological and aesthetic value (Prévosto et al. 2011). However, in abandoned grasslands, there may not always be significant invasion of shrubs and trees in secondary succession, even after 60 years (Bohner et al. 2019). It is assumed that the germination and establishment of woody plants is impeded by a virtually closed sward and by accumulated necromass, retarding further succession to these communities (Moog et al. 2002; Bohner et al. 2012). Concerning other attributes of plant growth forms, abandoned sites were characterised by greater coverage of geophytes than managed sites. This is probably because the plant storage organs are not damaged by trampling animals or oppressed by heavy mechanisms (e.g. mowing with tractors) at abandoned sites (McIntyre et al. 1995; Lavorel et al. 1999a) or simply because most of the geophytes possess the ability to regenerate vegetatively (Dupré and Diekmann 2001); i.e., they can use nutrients stored in rhizomes or bulbs to grow in the spring through the thick litter layer (Bobbink and Willems 1987).

With respect to life strategy, in our study, competitors prevailed at abandoned sites (see Rupprecht et al. 2016). Weiss and Jeltsch (2015) reported a strong increase in competitor plant types under resource (nutrient, water, and light)-rich conditions where the intensity of stress from their lack was minimal. The abandoned sites in our study were characterised by relatively high soil humidity (although not assessed exactly) and, at the same time, by the absence of disturbances, which favour long-lived plants.

Another evaluated functional trait was the reproduction type. Compared with those in the MGA, the average coverages of the species reproducing by seeds at the managed sites (MGM and MGP) were greater. Similarly, Rysiak et al. (2021) found an increasing number of species reproducing by seed on the managed plots. Seeds of numerous species can survive herbivore consumption or attach to fur, making herbivores vectors for plant dispersal (Malo and Suarez 1995; Cosyns et al. 2005). Also, increased light availability in managed sites can lead to better germination and seedling establishment (Jutila and Grace 2002).

Species reproducing by seeds possess persistent seed banks and their occurrence depends on the formation of bare ground and thus are expected to prefer grazed or mowed sites (e.g. Lavorel et al. 1999b). At the abandoned sites, species with the ability to reproduce vegetatively showed greater coverage values (cf. Pettit et al. 1995). The vegetative spread by rhizomes depends on a low frequency of disturbance (McIntyre and Lavorel 1994).

The results of our study showed that the managed sites allow the preservation of specific vegetation compositions with high coverage of graminoids. In contrast, abandoned sites are colonised primarily by perennial forbs (cf. Vannucchi et al. 2022). The high coverage values of short graminoids in managed grasslands were probably caused by better light conditions, with more opportunities to colonize the open space. Tall forbs increased their cover in abandoned meadows because, as strong competitors, they do not tolerate disturbance (Pavlů et al. 2011).

The effects of abandonment on soil properties, i.e., soil nitrogen and organic contents, confirm the key role of soil chemical properties in influencing vegetation in managed grasslands (Grime et al.,1997; White et al. 2004). Vegetation affects soil properties through a feedback mechanism (Petermann and Buzhdygan 2021). Some functional types, such as forbs, effectively increase nitrogen and carbon contents in grassland soils (Knops and Tilman 2000; De Deyn et al. 2009). Increased amounts of nitrogen and carbon in abandoned sites in comparison with managed sites (Vannucchi et al. 2022), which were also confirmed in our study, may be related to higher aboveground phytomass and increased plant litter decomposition (Gabarrón-Galeote et al. 2015; Bohner et al. 2019). Higher aboveground plant biomass and a denser surface layer of necromass at abandoned sites can also reduce the average soil temperature and increase soil moisture (Facelli and Pickett 1991).

Our study confirmed the differences in leaf persistence between abandoned and managed sites. Summer green species prevailed in coverage at the abandoned sites, whereas persistent green species prevailed at the grazing-managed sites. Rysiak et al. (2021) reported significantly greater coverage of summer green species at mowed sites. Leaf persistence is a functional trait that is important for plant competitiveness. It depends on the climate in the distribution range of the taxon and microclimate, as well as nutrient and light availability in typical habitats of the taxon.

We also observed differences in the PFT duration of flowering between abandoned and managed sites. Long-flowering species were characterised by great coverages at abandoned sites, whereas short-flowering species were dominant at mowed sites. At managed sites, this difference can be due to management timing (as well as intensity and frequency), which is important for the phenology of seed production (Poschlod et al. 2000). Plants in managed grassland biotopes are adapted to disturbances, e.g., by mechanisms related to palatability and mechanical defence (Callaway et al. 2000), growth form (Noy-Meir et al. 1989), and reproductive phenology (Lennartsson et al. 1998). Phenology can be considered particularly important in grazing, mown, and other managed biotopes because it determines whether a plant can produce seeds before the vegetation is disturbed (Akhalkatsi and Wagner 1996). However, in abandoned meadows, some species may experience longer or delayed development due to undecomposed litter, which acts as an insulating layer and fundamentally affects the physical and chemical properties of the soil (Janišová et al. 2007).

The similarities in the proportional coverage of the studied PFTs between the managed sites (MGM and MGP) were not surprising, as in both cases, applied management represents a disturbance to grassland habitats. In both types of management, mowing and grazing, aboveground vegetation is more or less regularly removed, but at least part of the phytomass is left (Kahmen and Poschlod 2008). This suggests that the effects of management on PFTs may operate mainly through the intensity of biomass loss or the biomass recovery rate, factors that were similar for both management practices in our study. Similar effects of mowing and grazing on PFTs indicate that different management practices may maintain communities with similar functional trait compositions and coverages, even though species compositions may differ (Moog et al. 2002; Garnier et al. 2007; Volf et al. 2016).

The grazed sites that we studied differed from the mowed ones mainly in terms of PFT plant height and seed mass. Small-height and small-seed species prevailed in the grazed sites and mowed sites hosted higher coverages of high-height and large-seed species. In contrast to mowing, grazing, as mentioned above, leads to the formation of gaps, as livestock trample and wallow within pastures (Gilhaus et al. 2017). Consequently, grazing promotes the establishment and maintenance of low-growth plants (Fleischer et al. 2013). Rysiak et al. (2021) noted that mowed and grazed sites are habitats rich in plants with relatively large seeds. Eriksson and Eriksson (1997) or Kahmen and Poschlod (2008) reported an increase in the number and coverage of species with small seeds caused by grazing. Seed mass is a functional trait related to plant dispersal and regeneration ability. Although heavier seeds may be dispersed over shorter distances than lighter seeds, heavier seeds may enhance seedling establishment, especially when light or nutrients are in short supply (Leishman et al. 2000). In terms of seed mass, seedlings from large seed species tend to survive better under a closed canopy (Grime et al. 1997), and an increase in the mean individual seed mass was reported (Westoby et al. 2002) to be a response trait associated with an increase in canopy height (Louault et al. 2005). Plants with large and heavy seeds have difficulty colonising isolated patches (Helsen et al. 2013), and their populations may become locally extinct unless they have good vegetative propagation ability or a long lifespan (Lindborg and Eriksson 2004; Bossuyt and Honnay 2006). In contrast, isolated patches may be more easily reached by plants with low seed mass (Westoby et al. 1996). A decline in the mean seed size in response to grazing was also reported by McIntyre and Lavorel (2001) for both perennial grasses and forbs. In our study, mowed meadows were characterized by greater canopy heights without isolated patches, especially in the time before mowing.

The authors have declared that no competing interests exist.

No ethical statement was reported.

This study was supported by the Scientific Grant Agency VEGA: Project No. 1/0076/22 “The Impact of Management on Biodiversity and Ecosystem Services of Submontane Meadows” and Project No. 1/0057/22 “Influence of environmental risk factors on phenological development of ecosystems in selected conservation areas of Slovakia“. This publication is the result of the project “Comprehensive research of deter­minants for ensuring environmental health” (ENVIHEALTH), ITMS 313011T721, supported by the Operational Programme Integrated Infrastructure (OPII) fund­ed by the European Regional Development Fund (ERDF).

Conceptualization: AD, SS. Data curation: HO, AD, MN. Formal analysis: AD, MN. Funding acquisition: SS. Investigation: AD, SS. Methodology: AD, MN. Project administration: AD. Writing - original draft: AD, MN, HO. Writing - review and editing: DV, SS, MN, AD, HO.

Andrea Diviaková https://orcid.org/0000-0002-3062-3976

Hana Ollerová https://orcid.org/0000-0002-7415-2192

Slavomír Stašiov https://orcid.org/0000-0002-4914-4465

Darina Veverková https://orcid.org/0000-0002-3421-1514

Milan Novikmec https://orcid.org/0000-0002-5192-4575

All of the data that support the findings of this study are available by authors upon request.